Related Experiment Video
Updated: Mar 15, 2026
07:24
Home-Based Monitor for Gait and Activity Analysis
Published on: August 8, 2019
7.4K
Ambulatory activity classification with dendogram-based support vector machine: Application in lower-limb active
Oishee Mazumder1, Ananda Sankar Kundu1, Prasanna Kumar Lenka2
1School of Mechatronics and Robotics, Indian Institute of Engineering Science and Technology, Shibpur, P.O. Botanic Garden, Shalimar, Howrah, West Bengal 711103, India.
Gait & Posture
|September 2, 2016
Summary
This study introduces a novel dendrogram-based support vector machine (DSVM) for classifying ambulatory activities in lower-limb exoskeleton users. The system achieves 95.2% accuracy, enabling smoother transitions for mobility assistance.
Area of Science:
- Biomedical Engineering
- Robotics
- Human-Computer Interaction
Background:
- Ambulatory activity classification is crucial for controlling mobility assistive devices like lower-limb exoskeletons.
- Current control methods often rely on manual switches or basic state machine logic, limiting natural interaction.
Purpose of the Study:
- To develop and evaluate a novel postural activity classifier for controlling lower-limb exoskeletons.
- To improve state transition accuracy and responsiveness in exoskeleton control systems.
Main Methods:
- Utilized a dendrogram-based support vector machine (DSVM) algorithm for classification.
- Integrated data from a pressure sensor-based wearable insole and two six-axis inertial measurement units (IMUs).
- Extracted polynomial coefficients from hip angle and center-of-pressure (CoP) trajectories as features for dynamic activity recognition.
Main Results:
- Successfully classified nine distinct ambulatory activities (2 static, 7 dynamic) with high accuracy.
- Achieved an overall classification accuracy of 95.2% for the proposed DSVM algorithm.
- Demonstrated the effectiveness of shape-based features from kinematic and pressure data.
Conclusions:
- The proposed DSVM-based activity classifier offers a robust and accurate method for controlling lower-limb exoskeletons.
- This approach enhances the potential for seamless and intuitive human-exoskeleton interaction.
- The findings pave the way for more sophisticated and responsive assistive device control.
